Recipient Organization
UNIV OF MARYLAND
(N/A)
COLLEGE PARK,MD 20742
Performing Department
(N/A)
Non Technical Summary
This research project addresses the BRAG program area of investigation and mitigation of unintended and off-target modifications by genome editing technologies. The cytosine base editors (CBEs), engineered from the CRISPR-Cas systems, are poised to revolutionize modern crop breeding due to their promising applications for targeted mutagenesis, allele replacement, and engineering of quantitative traits. CBEs can introduce premature stop codons to knock out genes and efficiently generate homozygous mutants within one generation. Hence, CBEs hold great promise for multiplexed editing in crops. It has been recently shown that multiplexed editing by Cas12a nucleases could lead to chromosomal rearrangements caused by concurrent DNA double-strand breaks (DSBs) in plants. Unlike Cas9 and Cas12a nucleases, CBEs derived from Cas9 and Cas12a generally do not produce DSBs in the plant genome. For example, Cas9-based CBEs mainly nick DNA, and Cas12a-based CBEs cannot cut DNA. Hence, it is anticipated that CBEs are unlikely to generate chromosomal rearrangements in a multiplexed editing setting. To investigate this important issue, we will first develop multiple improved Cas9-CBEs and Cas12a-CBEs for high-efficiency multiplexed editing in rice. Then we will conduct genome-wide off-target investigations in the rice plants with multiplexed edits using whole-genome sequencing. The relevance of the proposed project to the BRAG program is very significant because the results will provide critical information for the global regulatory agencies to come up with science-based regulatory policies as well as to guide regulatory decisions about the introduction of CRISPR-Cas generated organisms into the environment. The knowledge generated from our work will greatly aid plant scientists practicing gene editing in crops as well as facilitate the decision-making process at regulatory agencies such as the USDA and FDA.
Animal Health Component
40%
Research Effort Categories
Basic
60%
Applied
40%
Developmental
(N/A)
Goals / Objectives
Objective 1. Assess and identify top-performing Cas9-based CBEs for plant genome editing.In our preliminary results, we have identified many novel cytidine deaminases that displayed higher C-to-T editing efficiency than that of selected high-efficiency CBE controls. We will select the top eight cytidine deaminases from this collection and compare them with A3A/Y130F (high-activity cytidine deaminase with a broad editing window) and PmCDA1 (high-activity cytidine deaminase with a narrower window toward the 5' end with undetectable genome-wide off-target effects). These 10 cytidine deaminases will be configured in two CBE systems. The first CBE system is the standard BE3 or the V01 system. The second CBE system, V04, utilizes the MS2-MCP interaction to recruit multiple copies of UGI, which can drastically reduce the DSBs or indels at the target sites. Hence, we hypothesize that the CBE-V04 versions could help minimize DSB-related off-target effects (including chromosomal rearrangements) in a multiplexed editing setting. However, as some cytidine deaminases have shown decreased activity in the CBE-V04 configuration, we need to test all 8 novel cytidine deaminases so that we can identify the CBE-V04 editors that maintain high C-to-T editing efficiency with very minimal tendency to cause DNA DSBs. We will compare the 10 cytidine deaminases in these two CBE systems, first in rice protoplasts, and then in stable transgenic plants, to assess their genome editing efficiency and editing purity. With data from stable T0 lines, we will firmly benchmark our new CBEs (V01 and V04) against the commonly used A3A/Y130F CBEs and PmCDA1 CBEs. We expect to confirm at least four novel cytidine deaminases that, when configured in either the V01 or V04 version, will outperform or have similar performance to A3A/Y130F- and PmCDA1-based CBEs for on-target C-to-T editing efficiency. The editing windows by these different CBEs will be further confirmed and refined based on the genome editing data obtained in stable plants.Objective 2: Genome-wide investigation of off-target effects of top-performing Cas9-based CBEs in multiplexed editing in rice.We will select two top-performing cytidine deaminases, along with A3A/Y130F and PmCDA1, for genome-wide off-target analysis in multiplexed settings. Both CBE-V01 and V04 configurations will be compared. These eight CBEs will be compared in a lower-order multiplexed editing of four target sites by using our existing T-DNA vectors made earlier and a higher-order multiplexed editing of 16 target sites. We will choose the 16 target sites that are near or overlapping with the Cas12a target sites that we previously investigated for multiplexed editing by LbCas12a nuclease. This will allow us to compare our Cas9-based CBE data with our published Cas12a nuclease editing data and the Cas12a-based CBE data to be generated later in this project. The resulting T0 lines will be genotyped for genome editing efficiency at all target sites. Select edited lines will be used for whole genome sequencing to reveal potential off-target effects.With single nucleotide variations (SNVs) and indels revealed in replicates of three control types, we will be able to detect genomic variations between WT plants, any somaclonal variation that resulted from the tissue culture process, and any mutations resulting from the expression of CBEs in the lower- or higher-order multiplexed editing settings. We will identify all the potential off-target mutations and characterize them into sgRNA-dependent mutations (predictable based sequence similarity) and sgRNA-independent mutations (which would be the point of concern as they are caused by the cytidine deaminase alone). Among the two novel cytidine deaminases tested, we expect at least one would induce undetectable C-to-T off-target mutations, just like PmCDA1. Hence, our work would benchmark new CBEs that have high editing efficiency and minimal off-target effects, greatly enriching the CBE toolbox in plants. Because CBE-V01 editors can cause some levels of DNA DSBs, we expect to detect some chromosomal rearrangements, especially in those edited lines that endured higher-order multiplexed editing, as we revealed recently with Cas12a nuclease. However, we expect such chromosomal rearrangement events will decrease or disappear in the lines edited by CBE-V04 editors because these editors only render chromosomal nicking. With these data, we would have a clear picture of the different off-target effects of CBE-V01 and V04 editors in multiplexed editing, due to different DNA breaks and repair pathways involved.Objective 3: Assess and identify top-performing Cas12a-based CBEs for plant genome editing.Recently, we have developed LbCas12a-based CBEs by using an efficient LbCas12a variant (ttLbCas12a), an optimal linker, an efficient cytidine deaminase A3A/Y130F. Unfortunately, this LbCas12a CBE (called LbCas12a CBE-Verson-01/V01) is still not robust enough to work at every target site. Here, we aim to develop next-generation Cas12a CBEs with more robust editing activity by (1) exploring 8 new cytidine deaminases with high activity based on Objective 3.1 and (2) using the highly efficient LbCas12a-RRV variant. We call these new LbCas12a CBEs as LbCas12a-RRV version-02 (V02) editors. We will include A3A/Y130F in both LbCas12a CBE-V01 and V02 for comparison. As a result, a total of 10 LbCas12a CBEs will be generated by molecular cloning using the existing parts that are already in place (e.g., cytidine deaminases, linker, and deactivated LbCas12a variants).Like Cas9-based CBEs, we will test our LbCas12a-based CBEs in a lower-order multiplexed editing setting (with simultaneous editing of four target sites) and a higher-order multiplexed editing setting (with simultaneous editing of 16 or more target sites). With two multiplexed settings, we will generate a large amount of on-target editing data for 10 dLbCas12a CBEs at 19 (3+16) target sites with preferred TTTV PAMs and 12 (1+11) target sites with noncanonical TTV PAMs. These data, based on analysis of ~400 T0 plants (20 lines x 20 constructs) will help us to (1) identify the top-performing dLbCas12a CBEs, (2) reveal the editing windows of dLbCas12a CBEs, and (3) assess the performance of these CBEs at editing high-efficiency TTTV PAM sites vs. editing low-efficiency TTV PAM sites.Objective 4: Genome-wide investigation of off-target effects of top-performing Cas12a-based CBEs in multiplexed editing in rice.Based on our comprehensive on-target analysis of 10 different Cas12a CBEs in T0 lines, we will compare genome-wide off-target effects of four top-performing Cas12a CBEs in both multiplexed settings. We will use the edited plants identified in Objective 3 for WGS. For each construct, three independent T0 lines will be selected. Three GFP plants and three WT plants will be included as controls. In total, WGS (based on an Illumina platform) will be conducted for 24 edited plants (4 Cas12a CBEs x 2 multiplexed editing settings x 3 lines per case) and six control plants. We will follow the same WGS and data analysis pipeline that we have established.First, with WGS, we expect to validate all on-target editing that we confirmed by amplicon sequencing earlier. Second, we expect to detect very minimal gRNA-dependent off-target effects based on our previous study. Third, we anticipate detecting gRNA-independent off-target C-to-T base editing by certain high-activity cytidine deaminases, especially if such effects are detected with Cas9 CBEs. Fourth, we do not expect to detect any chromosomal rearrangement events by our Cas12a CBEs in both lower-order and higher-order multiplexed editing settings. This is because all the Cas12a CBEs tested in this study are based on dCas12a versions that cannot cut single-strand or double-strand DNA.
Project Methods
Objective 1. Assess and identify top-performing Cas9-based CBEs for plant genome editing.We will compare 8 top novel deaminases (BasA3G, MmA3AX1, EtA3C, CdA3G, OoA3GX2, DIA3G, LoA3GX1, and PsA3G1) identified in our preliminary results and two cytidine deaminase controls (A3A/Y130F and PmCDA1). The Gateway entry clones for the 10 CBE-V01 vectors have been made. We will generate 10 CBE-V04 entry vectors using Gibson Assembly to add the MCP-UGI extension, based on the CBE-V01 vectors. A T2A ribosome skipping sequence will be used to express two proteins via one transcript.To efficiently test these 20 CBEs (10 CBE-V01 editors and 10 CBE-V04 editors), we will generate multiplexed T-DNA vectors that allow for simultaneous editing of four relatively C-rich target sites in the rice genome. We will use our standard multiplexed CRISPR assembly approach to make these vectors via Golden Gate cloning and Gateway cloning. Briefly, the four protospacers will be synthesized as oligos, phosphorylated, and ligated into linearized pYPQ131C (with an OsU6 promoter), pYPQ132C (with an OsU6 promoter), pYPQ133D (with an OsU3 promoter), and pYPQ134D (with an OsU3 promoter). The resulting four vectors with cloned four sgRNAs will be used for Golden Gate cloning with pYPQ144, generating pYPQ144-sgRNAs Gateway entry clone. Then, a three-way Gateway reaction will be conducted, with each CBE Gateway entry clone, pYPQ144-sgRNAs Gateway entry clone, and pYPQ203 destination vector (with a ZmUbi promoter for driving CBE expression). The resulting 20 T-DNA vectors will be confirmed via plasmid NGS, and then used for rice protoplast transformation by following our established protocol. The PCR amplicon-based NGS will be conducted using a pooled sample strategy with the Hi-Tom barcoded primers. Sequencing data from the Illumina platform will be collected from three biological replicates of each construct. Data analysis will be done using software such as CRISPRmatch and CRISPResso2.To further benchmark the novel CBEs identified by the protoplast assay, we will analyze the chosen eight CBE-V01 editors (six novel CBE-V01 editors and two CBE-V01 control editors) and eight CBE-V04 editors (six novel CBE-V04 editors and two CBE-V04 control editors) in stable transgenic rice plants. For each construct, we will generate about 20 transgenic rice plants using Agrobacterium-mediated transformation by following our established protocol. The individual T0 lines will be genotyped by NGS with pooled samples using the Hi-Tom primers, followed up by a similar analysis pipeline that is used to analyze the protoplast editing data. The zygosity at target sites will be called by using our practiced criterion on monoallelic, biallelic, homozygous, and chimeric editing outcomes.Objective 2: Genome-wide investigation of off-target effects of top-performing Cas9-based CBEs in multiplexed editing in rice.The eight CBE-V01 editors, eight CBE-V04 editors, and one GFP tissue culture control will be used for rice transformation in the same batch. Three T0 lines per construct (17 constructs total), upon validation of editing at the target sites by amplicon-based NGS, will be selected for WGS, with three WT plants as controls. There will be a total of 54 (18x3) plants subjected to WGS. For each sample, we will target a sequencing depth of 50x. For the rice genome of ~400Mb, it will require an average of ~23Gb reads per sample. After high-quality genomic DNA is extracted from selected rice lines, DNA samples will be used for library preparation and high-throughput sequencing based on an Illumina NGS platform. The WGS data will be analyzed for off-target mutations according to our recent publications with the major steps illustrated in Figure 12. The potential chromosomal rearrangement events (e.g., large deletions, inversions, and translocations) will be further verified by PCR and Sanger sequencing.Objective 3: Assess and identify top-performing Cas12a-based CBEs for plant genome editing.We will target four previously used sites (two promoter sites in OsSWEET13 and OsSWEET14; two coding sequencing sites in OsGS3 and OsGW2) for the lower order multiplexed editing. The OsSWEET14 site contains a non-canonical TTV PAM which can be edited by LbCas12a, albeit with low efficiency, and editing efficiency at TTV PAM sites could be augmented by the LbCas12a-RRV variant. For the higher-order editing, the 16 crRNAs targeting 16 TTTV PAM sites that we used previously will be chosen. Because 11 out of these 16 crRNAs can each target a second site with the same protospacer sequence but relaxed TTV PAM requirement, these 16 crRNAs can presumably target up to 27 (16+11) target sites in the rice genome, presenting a unique case to study highly multiplexed cytosine base editing at sites with different editing activities.In total, we will generate 20 dLbCas12a-based CBEs, 10 for multiplexed editing of four target sites and 10 for multiplexed editing of 16 to 27 sites in rice. The multiplexed vectors will be assembled based on our dual Pol II and dual ribozyme processing system, and confirmed by plasmid NGS. Due to the intrinsic low activity of dCas12a CBEs in protoplasts, we will directly compare these vectors in stable rice lines for investigation of on-target and off-target effects. These 20 constructs, along with the GFP control construct, will be used for Agrobacterium-mediated rice transformation. For each Cas12a CBE construct, about 20 T0 lines will be analyzed by following the same NGS protocol employed earlier.Objective 4: Genome-wide investigation of off-target effects of top-performing Cas12a-based CBEs in multiplexed editing in rice.Based on our comprehensive on-target analysis of 10 different Cas12a CBEs in T0 lines, we will compare genome-wide off-target effects of four top-performing Cas12a CBEs in both multiplexed settings. We will use the edited plants identified in Objective 3 for WGS. For each construct, three independent T0 lines will be selected. Three GFP plants and three WT plants will be included as controls. In total, WGS (based on an Illumina platform) will be conducted for 24 edited plants (4 Cas12a CBEs x 2 multiplexed editing settings x 3 lines per case) and six control plants. For each sample, we will target a sequencing depth of 50x. For the rice genome of ~400Mb, we will require an average of ~23Gb reads per sample. After high-quality genomic DNA is extracted from selected rice lines, DNA samples will be used for library preparation and high-throughput sequencing based on an Illumina HiSeq or similar platform. The WGS data will be analyzed for off-target mutations according to our established pipeline. Briefly, adapters of raw sequencing reads will be trimmed using SKEWER tool and the Illumina TruSeq adapter. Cleaned reads from the rice will be mapped to the reference sequence of rice (http://rice.plantbiology.msu.edu/).The Genome Analysis Toolkit (GATK) will be used to realign reads near indels and recalibrate base quality scores. A known single nucleotide polymorphism (SNP) and insertion/deletion (indel) database for GATK best practices will be downloaded for rice from the SNP-Seek Database (http://snp-seek.irri.org/). Whole genome SNPs or indels will be detected with three independent software programs. These mutations will be mapped to putative off-target sites that are computationally predicted, allowing for the identification of true off-target mutations.